松毛虫危害下的马尾松林冠层光谱特征可辨性分析  被引量:6

Canopy spectral characteristics distinguishability analysis of Pinus massoniana forests with Dendrolimus punctatus Walker damage

在线阅读下载全文

作  者:许章华[1] 刘健[2,3,4] 陈崇成[5] 余坤勇[2,3] 黄旭影 王美雅 

机构地区:[1]福州大学环境与资源学院,福州350116 [2]福建农林大学3S技术应用研究所,福州350002 [3]福建农林大学林学院,福州350002 [4]三明学院,三明365000 [5]福州大学地理空间信息技术国家地方联合工程研究中心,福州350002

出  处:《国土资源遥感》2016年第2期41-47,共7页Remote Sensing for Land & Resources

基  金:国家自然科学基金青年科学基金项目(编号:41501361);福建省资源环境监测与可持续经营利用重点实验室开放基金项目(编号:ZD1403);福州大学人才基金项目(编号:XRC-1345)共同资助

摘  要:深入挖掘寄主光谱响应机制是推进马尾松毛虫害遥感快速监测与预警的必要基础。将采集于福建省长汀县、南平市建阳区的46条马尾松林冠层光谱曲线数据设为规则组,利用单因素方差分析法实现不同危害等级可辨性波长的选择。研究表明,不同危害等级下的马尾松林冠层光谱数据呈现极显著差异(P〈0.01),其中,中度—重度危害的马尾松林冠层光谱可辨性在516.51~598.99 nm和700.68~706.18 nm位置上有显著差异(P〈0.05),在708.92~810.62 nm位置上有极显著差异(P〈0.01)。为此,以519.20 nm,540.72 nm,758.40 nm和785.88 nm波段处光谱反射率为组合,以健康状态下的马尾松林冠层光谱数据为标准样本,基于空间距离法、相关系数法及光谱角制图法分别建立松毛虫危害等级的定量化判定规则,并利用将乐县、南平市延平区、华安县的34条验证组光谱曲线数据对此规则进行验证。结果显示,空间距离法的判定效果远优于相关系数法与光谱角制图法;无松毛虫危害、轻度危害、中度危害以及重度危害的空间距离判定规则依次为:〈0.355 3,[0.355 3,0.742 5),[0.742 5,0.9631)及≥0.963 1,判定精度为88.24%,准确率达97.06%。The deep mining of host spectral response mechanism is the necessary foundation for Dendrolimus punctatus Walker damage remote sensing fast monitoring and early warning. 46 canopy spectral curve data of Pinus massoniana forests collected in Changting County of Jianyang District were set as the rule group,and the one- way ANOVA was used to realize the distinguished wavelengths selection with different pest levels,and the results showed that there were highly significant differences of pine forests canopy spectral data with different pest levels( P〈0. 01),in which there were significant differences at 516. 51 ~ 598. 99 nm and 700. 68 ~ 706. 18 nm of spectral distinguishability between moderate damage and severe damage( P〈0. 05),and highly significant differences at708. 92 ~ 810. 62 nm( P〈0. 01). Thus,based on the combination of spectral reflectance of 519. 2 nm,540. 72 nm,758. 4 nm,785. 88 nm and taking the healthy pine forests canopy spectral data as the standard sample,the authors constructed the quantitative determination rules of pest levels of Dendrolimus punctatus Walker,relying on the methods of spatial distance,correlation coefficient and spectral angle mapping respectively. The rules were verified with the test group of 34 spectral curve data collected in Jiangle County,Yanping District of Nanping City,and Huaan County,and the results showed that the determination effect of spatial distance method was by far better than that of the correlation coefficient method and spectral angle mapping method. The spatial distance determination rules of non- damage,mild damage,moderate damage and severe Dendrolimus punctatus Walker damages were as follows: 0. 355 3,[0. 355 3,0. 742 5),[0. 742 5,0. 963 1) and ≥ 0. 9631,with the determination accuracy being 88. 24% and the accurate rate being 97. 06%.

关 键 词:马尾松毛虫害 光谱特征可辨性 单因素方差分析 空间距离法 相关系数法 光谱角制图法 

分 类 号:TP79[自动化与计算机技术—检测技术与自动化装置] P237[自动化与计算机技术—控制科学与工程]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象